Because fracture of nickel-titanium rotary endodontic instruments sometimes occurs clinically, there is considerable interest in failure mechanisms. One possible mechanism is that excessive work hardening during clinical use leads to instrument fracture. Purpose: Employ measurements of Vickers hardness to investigate work hardening in clinically used rotary instruments, since values of Vickers hardness have been reported (Brantley, Orthodontic Materials: Scientific and Clinical Aspects, 2001; AADR Abst. No. 24, 1998) for nickel-titanium orthodontic wires having similar compositions and properties. Methods: A variety of nickel-titanium rotary instruments were obtained with unknown number of clinical uses and sterilization history. Some instruments had fractured; others had experienced considerable permanent torsional deformation; and the remainder had no visual evidence of damage. Instruments were resin-mounted and sectioned with a low-speed, water-cooled diamond saw. After polishing the test specimens with 0.05 mm alumina, Vickers hardness was measured near the flutes at D2 and between approximately D10 and D15, and beyond D16, for nine representative used instruments (N = 5 for each region), with a 300 gm load. Consistent hardness values could not be obtained at D1. Results: Vickers hardness varied at different locations on the used instruments. Maximum values up to 481 ± 31 (mean ± SD) were found beyond D16, and minimum values occurred at D2 to D4, where means ranged from 313 to 324 (SD from 7 to 16). Vickers hardness at D2 to D4 was 320 ± 7 for an instrument serving as a control. Similar mean Vickers hardness values from 286 to 332 (SD from 5 to 19) were previously reported for shape-memory nickel-titanium orthodontic wires. Conclusions: The rotary instruments did not experience substantial work hardening at D2 to D4 during clinical use, which is consistent with results from recent differential scanning calorimetric analyses (IADR Abst No. 3854, 2002).